CMPU4091 Visualizing Data

options(repos = c(CRAN = "https://cloud.r-project.org"))

# Function to quietly install and load packages
quiet_library <- function(pkg) {
  suppressMessages(suppressWarnings({
    if (!requireNamespace(pkg, quietly = TRUE)) {
      install.packages(pkg, quiet = TRUE) # Install without messages
    }
    library(pkg, character.only = TRUE) # Load silently
  }))
}

# Install & Load Packages Silently
quiet_library("ggplot2")
quiet_library("plotly")
quiet_library("gapminder") #An excerpt of the data available at Gapminder.org. For each of 142 countries, the package provides values for life expectancy, GDP per capita, and population, every five years, from 1952 to 2007.

package 'gapminder' successfully unpacked and MD5 sums checked

quiet_library("tidyverse")
quiet_library("ggstream") #For Stream graphs

package 'ggstream' successfully unpacked and MD5 sums checked

quiet_library("fmsb")  # For radar chart

package 'fmsb' successfully unpacked and MD5 sums checked

quiet_library("GGally")  # For parallel co-ordinates chart

1. Time Trends

1.1 Bar Charts

1.1.1 Simple Time Bar Chart

p_bar <- gapminder %>%
  group_by(year, continent) %>%
  summarize(lifeExp = mean(lifeExp), .groups = "drop") %>%  # Aggregate life expectancy
  ggplot(aes(x = factor(year), y = lifeExp, fill = continent)) +  # Convert year to factor for discrete bars
  geom_col(position = "dodge") +  # Use geom_col() for summarized data
  labs(title = "Average Life Expectancy Over Time by Continent",
       x = "Year",
       y = "Life Expectancy",
       fill = "Continent") +
  theme_bw()

p_bar

1.1.2 Interactive Simple Time Bar Chart (using ggplotly)

#Making it interactive (check the hover text, and click on the legend to add/remove bars)
ggplotly(p_bar)

1.1.3 Divergent Bar Chart

# Divergent bar chart
p_bar_divergent <- gapminder %>%
  group_by(year) %>%
  mutate(global_avg = mean(lifeExp)) %>%  # Compute global average per year
  ungroup() %>%
  group_by(year, continent) %>%
  summarize(lifeExp = mean(lifeExp), global_avg = unique(global_avg), .groups = "drop") %>%
  mutate(deviation = lifeExp - global_avg) %>%  # Calculate deviation from the global average
  ggplot(aes(x = factor(year), y = deviation, fill = continent)) +
  geom_col(position = "dodge") +  # Dodge to show bars side by side
  geom_hline(yintercept = 0, color = "black", linetype = "dashed") +  # Add a baseline at 0
  labs(title = "Divergent Bar Chart of Life Expectancy by Continent",
       x = "Year",
       y = "Deviation from Global Average Life Expectancy",
       fill = "Continent") +
  theme_bw()

p_bar_divergent

1.1.4 Divergent Bar Chart Vertical Reference line

# Prepare data (USArrests is a baseline r dataset)
state_crime <- USArrests %>%
  mutate(State = rownames(USArrests)) %>%  # Convert row names to a column
  mutate(Deviation = Murder - mean(Murder)) %>%  # Calculate deviation from national average
  arrange(Deviation)  # Order by deviation for better visualization

# Create Divergent Bar Chart
#Shows States with higher-than-average murder rates (bars extending right, colored blue).
#States with lower-than-average murder rates (bars extending left, colored red).
#The vertical dashed line at 0 makes it easy to compare states.

p <- ggplot(state_crime, aes(y = reorder(State, Deviation), x = Deviation, fill = Deviation > 0)) +
  geom_col(width = 0.7, position = position_dodge(0.9)) + # Bar Chart set bar width and separation
  geom_vline(xintercept = 0, color = "black", linetype = "dashed") +  # Vertical reference line
  scale_fill_manual(values = c("red", "blue"), guide = "none") +  # Color: Red = below avg, Blue = above avg
  labs(title = "Deviation of Murder Rate from National Average (US States)",
       x = "Deviation from National Average",
       y = "State") +
  theme_minimal()

1.1.5 Interactive Divergent Bar Chart (with Defined Tooltips ggplotly)

# Interactive Chart

# Add tool tips
# Create ggplot with tooltips
p <- ggplot(state_crime, aes(y = reorder(State, Deviation), x = Deviation, fill = Deviation > 0, 
                             text = paste0("State: ", State, "<br>",
                                           "Deviation: ", round(Deviation, 2), "<br>",
                                           "Murder Rate: ", round(Murder, 2)))) + 
  geom_col(width = 0.7, position = position_dodge(0.9)) +  # Set bar width and separation
  geom_vline(xintercept = 0, color = "black", linetype = "dashed") +  # Vertical reference line
  scale_fill_manual(values = c("red", "blue"), guide = "none") +  # Color: Red = below avg, Blue = above avg
  labs(title = "Deviation of Murder Rate from National Average (US States) with tooltips",
       x = "Deviation from National Average",
       y = "State") +
  theme_minimal()


# Convert to interactive plot with tooltips
ggplotly(p, tooltip = "text")

1.2 Line Charts

1.2.1 Single Line Time Trend

# Single Time Trend - Life Expectancy in the United States
p_single <- ggplot(gapminder %>% filter(country == "United States"), aes(x = year, y = lifeExp)) +
  geom_line() +
  labs(title = "Life Expectancy in the United States", x = "Year", y = "Life Expectancy")
ggplotly(p_single)

1.2.2 Time Trend Faceted

# Faceting (Life expectancy by continent)
p_faceted <- gapminder %>%
  ggplot(aes(x = year, y = lifeExp, color = continent)) +
  geom_line() +
  facet_wrap(~ continent, ncol = 3) +  # Facet by continent, 3 columns
  labs(title = "Life Expectancy Over Time by Continent", x = "Year", y = "Life Expectancy", color = "Continent") +
  theme_bw()

ggplotly(p_faceted)

1.2.3 Interactive Step Chart (using ggplotly)

# Filter data for the United States
p_step <- ggplot(gapminder %>% filter(country == "United States"), aes(x = year, y = lifeExp)) +
  geom_step(linewidth = 1, color = "blue") +  # Step chart for discrete changes
  geom_point(size = 3, color = "red") +  # Highlight actual data points
  labs(
    title = "Step Chart: Life Expectancy in the United States",
    x = "Year",
    y = "Life Expectancy"
  ) +
  theme_minimal()

# Convert to interactive Plotly chart
ggplotly(p_step)

1.2.4 Multiple Line Chart

# Get a list of European countries (filter from gapminder dataset where continent="Europe")

european_countries <- unique(gapminder$country[gapminder$continent == "Europe"])

# Multiple Time Trends (Life expectancy for European countries)
p_multiple_europe <- gapminder %>%
  filter(country %in% european_countries) %>%
  ggplot(aes(x = year, y = lifeExp, color = country)) +
  geom_line() +
  scale_color_viridis_d(option = "plasma") +
  labs(title = "Life Expectancy Over Time in European Countries", x = "Year", y = "Life Expectancy", color = "Country") +
  theme_bw()
#scale_color_viridis_d(option = "plasma") set the colours smoothly

1.2.5 Interactive Multiple Line Chart (ggplotly)

# Make it interactive
ggplotly(p_multiple_europe)

1.3 Area Charts

1.3.1 Stacked Area Chart

# Summarize population by year and continent
gapminder_summary <- gapminder %>%
  group_by(year, continent) %>%
  summarize(population = sum(pop), .groups = "drop")  # Total population per continent per year

# Create the stacked area chart
areagraph <- ggplot(data = gapminder_summary, 
                    aes(x = year, 
                        y = population, 
                        fill = continent, 
                        group = continent, 
                        text = paste("<b>Continent:</b> ", continent, 
                                     "<br><b>Year:</b> ", year, 
                                     "<br><b>Population:</b> ", scales::comma(population)))) +
  geom_area(position = "stack", alpha = 0.8) +  # Stacked area chart
  scale_fill_brewer(palette = "Set3") +  # Nice color palette
  scale_y_continuous(labels = scales::comma) +  # Format population numbers
  labs(
    title = "Population Trends by Continent (1952-2007)",
    x = "Year",
    y = "Population",
    fill = "Continent"
  ) +
  theme_minimal()

# Convert to interactive Plotly chart
ggplotly(areagraph, tooltip = "text")

1.3.2 Streamgraph

# Aggregate population by year and continent
gapminder_stream <- gapminder %>%
  group_by(year, continent) %>%
  summarise(pop = sum(pop), .groups = "drop")  # Sum population per continent per year

# Create the streamgraph
streamgraph <- ggplot(data = gapminder_stream, 
                      aes(x = year, 
                          y = pop, 
                          fill = continent,  # Fill by continent for distinct regions
                          group = continent, 
                          text = paste("Continent: ", continent, 
                                       "<br>Year: ", year, 
                                       "<br>Population: ", format(pop, big.mark = ",")))) +  # Format population
  geom_stream(type = "proportional", alpha = 0.8) +  # Proper streamgraph
  scale_fill_brewer(palette = "Set3") +  # Nice color palette
  scale_y_continuous(expand = c(0, 0)) +  # Remove extra space
  labs(
    title = "World Population Trends by Continent (1952-2007)",
    x = "Year",
    y = NULL,  # Streamgraphs don't need y-axis labels
    fill = "Continent"
  ) +
  theme_minimal() + 
  theme(axis.text.y = element_blank(),  # Remove y-axis text
        axis.ticks.y = element_blank())

# Convert to interactive Plotly chart
ggplotly(streamgraph, tooltip = "text")

1.3.3 Streamgraph - Ridged

# Create the streamgraph
streamgraph <- ggplot(data = gapminder_stream, 
                      aes(x = year, 
                          y = pop, 
                          fill = continent,  # Fill by continent for distinct regions
                          group = continent, 
                          text = paste("Continent: ", continent, 
                                       "<br>Year: ", year, 
                                       "<br>Population: ", format(pop, big.mark = ",")))) +  # Format population
  geom_stream(type = "ridge", alpha = 0.8) +  # Proper streamgraph
  scale_fill_brewer(palette = "Set3") +  # Nice color palette
  scale_y_continuous(expand = c(0, 0)) +  # Remove extra space
  labs(
    title = "World Population Trends by Continent (1952-2007)",
    x = "Year",
    y = NULL,  # Streamgraphs don't need y-axis labels
    fill = "Continent"
  ) +
  theme_minimal() + 
  theme(axis.text.y = element_blank(),  # Remove y-axis text
        axis.ticks.y = element_blank())

# Convert to interactive Plotly chart
ggplotly(streamgraph, tooltip = "text")

1.3.4 Radar chart

# Filter gapminder data for 2007 and group by continent
gapminder_radar <- gapminder %>%
  filter(year == 2007) %>%
  group_by(continent) %>%
  summarise(
    lifeExp = mean(lifeExp),  # Average life expectancy
    gdpPercap = mean(gdpPercap),  # Average GDP per capita
    pop = mean(pop),  # Average population
    .groups = "drop"
  )

# Normalize the values (Radar charts require similar scales)
normalize <- function(x) (x - min(x)) / (max(x) - min(x))

gapminder_radar[, 2:4] <- lapply(gapminder_radar[, 2:4], normalize)

# Add max/min rows required for fmsb package
radar_data <- rbind(
  rep(1, ncol(gapminder_radar)-1),  # Max value (normalized)
  rep(0, ncol(gapminder_radar)-1),  # Min value (normalized)
  gapminder_radar[, -1]  # Actual data
)

# Set row names to continent names
rownames(radar_data) <- c("Max", "Min", gapminder_radar$continent)

Warning: Setting row names on a tibble is deprecated.

# Plot the radar chart
radarchart(radar_data, axistype = 1, 
           pcol = c("red", "blue", "green", "purple", "orange"),  # Colors for each continent
           pfcol = alpha(c("red", "blue", "green", "purple", "orange"), 0.4), 
           plwd = 2, 
           title = "Radar Chart: Life Expectancy, GDP per Capita, and Population (2007)")

# Add legend
legend("topright", legend = gapminder_radar$continent, 
       col = c("red", "blue", "green", "purple", "orange"), 
       pch = 16, bty = "n")

1.3.5 Circular Bar Chart

# Prepare the data (normalize values)
gapminder_circular <- gapminder %>%
  filter(year == 2007) %>%
  group_by(continent) %>%
  summarise(
    lifeExp = mean(lifeExp),
    gdpPercap = mean(gdpPercap),
    pop = mean(pop),
    .groups = "drop"
  ) %>%
  mutate(across(lifeExp:pop, ~ (. - min(.)) / (max(.) - min(.)))) %>%  # Normalize values
  pivot_longer(cols = -continent, names_to = "Metric", values_to = "Value") %>%
  arrange(Metric, desc(Value))  # Arrange for a better circular order

# Set factor levels for better ordering
gapminder_circular$continent <- factor(gapminder_circular$continent, levels = rev(unique(gapminder_circular$continent)))

# Plot Circular Bar Chart
ggplot(gapminder_circular, aes(x = continent, y = Value, fill = Metric)) +
  geom_bar(stat = "identity") +
  coord_polar(start = 0) +  # Convert to circular layout
  facet_wrap(~ Metric, ncol = 1) +  # Separate for each metric
  theme_minimal() +
  labs(title = "Circular Bar Chart: Life Expectancy, GDP, and Population (2007)",
       x = "", y = "") +
  theme(axis.text.x = element_blank(), axis.ticks = element_blank())

1.3.6 Faceted bar chart

# Define normalization function
normalize <- function(x) (x - min(x)) / (max(x) - min(x))

gapminder_bar <- gapminder %>%
  filter(year == 2007) %>%
  group_by(continent) %>%
  summarise(
    lifeExp = mean(lifeExp),
    gdpPercap = mean(gdpPercap),
    pop = mean(pop),
    .groups = "drop"
  ) %>%
  mutate(across(lifeExp:pop, normalize)) %>%
  pivot_longer(cols = -continent, names_to = "Metric", values_to = "Value")

# Plot
ggplot(gapminder_bar, aes(x = continent, y = Value, fill = continent)) +
  geom_bar(stat = "identity", position = "dodge") +
  facet_wrap(~ Metric, scales = "free_y") +
  labs(title = "Faceted Bar Charts: Life Expectancy, GDP, and Population (2007)") +
  theme_minimal()

1.3.7 Parallel Co-ordinates Chart

library(ggplot2)
library(GGally)

gapminder_parallel <- gapminder %>%
  filter(year == 2007) %>%
  group_by(continent) %>%
  summarise(
    lifeExp = mean(lifeExp),
    gdpPercap = mean(gdpPercap),
    pop = mean(pop),
    .groups = "drop"
  ) %>%
  mutate(across(lifeExp:pop, normalize))  # Normalize data

# Plot
ggparcoord(gapminder_parallel, columns = 2:4, groupColumn = 1, scale = "globalminmax") +
  labs(title = "Parallel Coordinates Plot: Life Expectancy, GDP, and Population (2007)") +
  theme_minimal()